From: Luis R. Rodriguez <lrodriguez@atheros.com>
To: ath9k-devel@lists.ath9k.org
Subject: [ath9k-devel] throughput is unstable
Date: Mon, 4 Jan 2010 07:55:04 -0800 [thread overview]
Message-ID: <20100104155504.GC3855@tux> (raw)
In-Reply-To: <40101cc30912301258s6f70d39ei7f03dc5013d758a4@mail.gmail.com>
On Wed, Dec 30, 2009 at 12:58:51PM -0800, rootkit85 at yahoo.it wrote:
> 2009/12/29 Bj?rn Smedman <bjorn.smedman@venatech.se>:
> > Nope, haven't solved it. I think I've patched the driver in 50 places
> > and it still goes up and down up and down in perfect cycles. Who needs
> > an atomic clock when you can just count the throughput cycles and know
> > exactly what time it is? :)
> >
> > My current theory is that it's some sort of tx beamforming algo that
> > screws up once signal gets below -78 dBm or so. But you see the stalls
> > even with perfect signal, right? That never happens to me (which is
> > fortunate as I would be in an asylum by now if it did).
> >
> > Does anybody know which register to fiddle with to turn off all
> > adaptive tx stuff (to debug)? It's an AR5416 MAC/BB with AR2133 radio.
> > Is it possible to get some hw documentation somewhere?
> >
> > BTW, my setup is an unencrypted HT20 access point.
> >
> > /Bj?rn
> >
> > 2009/12/26 <rootkit85@yahoo.it>:
> >> 2009/12/24 Bj?rn Smedman <bjorn.smedman@venatech.se>:
> >>> Do the stalls occur at regular intervals? I have a similar problem on
> >>> an AR9103 in AP mode. See my previous post 'Cyclic throughput at a
> >>> distance'.
> >>>
> >>> /Bj?rn
> >>>
> >>> On Thu, Dec 24, 2009 at 2:31 PM, <rootkit85@yahoo.it> wrote:
> >>>> On Thu, Dec 24, 2009 at 2:27 PM, <rootkit85@yahoo.it> wrote:
> >>>>> Hi,
> >>>>> I'm using an AR9223 in AP mode with latest compat-wireless-2009-12-11
> >>>>> drivers but I get bad performances.
> >>>>> SSH is unusable.
> >>>>> The link is fine when it works but sometimes it stalls for a few
> >>>>> seconds, which kills tcp performances:
> >>>>>
> >>>>> iperf -sui1
> >>>>> ------------------------------------------------------------
> >>>>> Server listening on UDP port 5001
> >>>>> Receiving 1470 byte datagrams
> >>>>> UDP buffer size: 106 KByte (default)
> >>>>> ------------------------------------------------------------
> >>>>> [ 3] local 192.168.0.1 port 5001 connected with 192.168.0.79 port 33066
> >>>>> [ ID] Interval Transfer Bandwidth Jitter Lost/Total Datagrams
> >>>>> [ 3] 0.0- 1.0 sec 7.98 MBytes 66.9 Mbits/sec 0.117 ms 1423/ 7115 (20%)
> >>>>> [ 3] 1.0- 2.0 sec 11.3 MBytes 94.5 Mbits/sec 0.085 ms 2811/10849 (26%)
> >>>>> [ 3] 2.0- 3.0 sec 11.1 MBytes 92.9 Mbits/sec 0.119 ms 2734/10632 (26%)
> >>>>> [ 3] 3.0- 4.0 sec 6.63 MBytes 55.6 Mbits/sec 0.135 ms 1257/ 5986 (21%)
> >>>>> [ 3] 4.0- 5.0 sec 1.95 MBytes 16.3 Mbits/sec 0.113 ms 186/ 1575 (12%)
> >>>>> [ 3] 5.0- 6.0 sec 1.17 MBytes 9.81 Mbits/sec 0.083 ms 254/ 1088 (23%)
> >>>>> [ 3] 6.0- 7.0 sec 0.00 Bytes 0.00 bits/sec 0.083 ms 0/ 0 (nan%)
> >>>>> [ 3] 7.0- 8.0 sec 0.00 Bytes 0.00 bits/sec 0.083 ms 0/ 0 (nan%)
> >>>>> [ 3] 8.0- 9.0 sec 0.00 Bytes 0.00 bits/sec 0.083 ms 0/ 0 (nan%)
> >>>>> [ 3] 9.0-10.0 sec 1.86 MBytes 15.6 Mbits/sec 0.144 ms 649/ 1973 (33%)
> >>>>> [ 3] 10.0-11.0 sec 2.38 MBytes 19.9 Mbits/sec 0.097 ms 609/ 2304 (26%)
> >>>>> [ 3] 11.0-12.0 sec 652 KBytes 5.34 Mbits/sec 0.068 ms 198/ 652 (30%)
> >>>>> [ 3] 12.0-13.0 sec 9.69 MBytes 81.3 Mbits/sec 0.085 ms 2563/ 9476 (27%)
> >>>>> [ 3] 13.0-14.0 sec 6.38 MBytes 53.5 Mbits/sec 0.249 ms 1426/ 5977 (24%)
> >>>>> [ 3] 14.0-15.0 sec 8.45 MBytes 70.9 Mbits/sec 1.677 ms 1940/ 7971 (24%)
> >>>>> [ 3] 15.0-16.0 sec 8.03 MBytes 67.4 Mbits/sec 2.580 ms 1899/ 7628 (25%)
> >>>>> [ 3] 16.0-17.0 sec 6.91 MBytes 57.9 Mbits/sec 0.157 ms 905/ 5832 (16%)
> >>>>> [ 3] 17.0-18.0 sec 233 KBytes 1.91 Mbits/sec 0.788 ms 32/ 194 (16%)
> >>>>> [ 3] 18.0-19.0 sec 1.12 MBytes 9.38 Mbits/sec 0.620 ms 210/ 1008 (21%)
> >>>>> [ 3] 19.0-20.0 sec 7.06 MBytes 59.3 Mbits/sec 0.159 ms 1027/ 6066 (17%)
> >>>>> [ 3] 20.0-21.0 sec 6.07 MBytes 50.9 Mbits/sec 0.322 ms 720/ 5047 (14%)
> >>>>> [ 3] 21.0-22.0 sec 6.13 MBytes 51.4 Mbits/sec 0.109 ms 619/ 4989 (12%)
> >>>>> [ 3] 22.0-23.0 sec 4.84 MBytes 40.6 Mbits/sec 0.154 ms 365/ 3814 (9.6%)
> >>>>> [ 3] 23.0-24.0 sec 5.26 MBytes 44.1 Mbits/sec 0.308 ms 224/ 3978 (5.6%)
> >>>>> [ 3] 24.0-25.0 sec 6.06 MBytes 50.9 Mbits/sec 0.167 ms 757/ 5083 (15%)
> >>>>> [ 3] 25.0-26.0 sec 6.84 MBytes 57.4 Mbits/sec 0.148 ms 902/ 5779 (16%)
> >>>>> [ 3] 26.0-27.0 sec 7.66 MBytes 64.3 Mbits/sec 0.152 ms 1288/ 6755 (19%)
> >>>>> [ 3] 27.0-28.0 sec 0.00 Bytes 0.00 bits/sec 0.152 ms 0/ 0 (nan%)
> >>>>> [ 3] 28.0-29.0 sec 84.7 KBytes 694 Kbits/sec 14.385 ms 141/ 200 (70%)
> >>>>> [ 3] 0.0-30.0 sec 145 MBytes 40.7 Mbits/sec 0.290 ms 27599/131272 (21%)
> >>>>> [ 3] 0.0-30.0 sec 1 datagrams received out-of-order
> >>>>>
> >>>>>
> >>>>> $ ping 192.168.0.1
> >>>>> PING 192.168.0.1 (192.168.0.1) 56(84) bytes of data.
> >>>>>
> >>>>> 64 bytes from 192.168.0.1: icmp_seq=2 ttl=64 time=280 ms
> >>>>> 64 bytes from 192.168.0.1: icmp_seq=3 ttl=64 time=1.95 ms
> >>>>> 64 bytes from 192.168.0.1: icmp_seq=4 ttl=64 time=1002 ms
> >>>>> 64 bytes from 192.168.0.1: icmp_seq=5 ttl=64 time=50.3 ms
> >>>>> 64 bytes from 192.168.0.1: icmp_seq=6 ttl=64 time=264 ms
> >>>>> 64 bytes from 192.168.0.1: icmp_seq=7 ttl=64 time=322 ms
> >>>>> 64 bytes from 192.168.0.1: icmp_seq=8 ttl=64 time=1232 ms
> >>>>> 64 bytes from 192.168.0.1: icmp_seq=10 ttl=64 time=5.11 ms
> >>>>> 64 bytes from 192.168.0.1: icmp_seq=11 ttl=64 time=1.96 ms
> >>>>> ^C
> >>>>> --- 192.168.0.1 ping statistics ---
> >>>>> 11 packets transmitted, 9 received, 18% packet loss, time 10021ms
> >>>>> rtt min/avg/max/mdev = 1.957/351.321/1232.350/430.558 ms, pipe 2
> >>>>>
> >>>>>
> >>>>> there are no errors in the master and client syslog, all the clients
> >>>>> are using the same ath9k with an AR5416 card.
> >>>>> signal level is around -61 dBm at 3 meters, without walls.
> >>>>>
> >>>>
> >>>> This is an iperf from the AP to the client, definitely worse:
> >>>>
> >>>> ------------------------------------------------------------
> >>>> Server listening on UDP port 5001
> >>>> Receiving 1470 byte datagrams
> >>>> UDP buffer size: 112 KByte (default)
> >>>> ------------------------------------------------------------
> >>>> [ 3] local 192.168.0.79 port 5001 connected with 192.168.0.1 port 59087
> >>>> [ ID] Interval Transfer Bandwidth Jitter Lost/Total Datagrams
> >>>> [ 3] 0.0- 1.0 sec 159 KBytes 1.31 Mbits/sec 9.106 ms 100/ 211 (47%)
> >>>> [ 3] 1.0- 2.0 sec 2.83 MBytes 23.7 Mbits/sec 14.785 ms 221/ 2240 (9.9%)
> >>>> [ 3] 2.0- 3.0 sec 876 KBytes 7.17 Mbits/sec 0.210 ms 45/ 655 (6.9%)
> >>>> [ 3] 3.0- 4.0 sec 2.29 MBytes 19.2 Mbits/sec 0.137 ms 122/ 1753 (7%)
> >>>> [ 3] 4.0- 5.0 sec 4.36 MBytes 36.6 Mbits/sec 0.663 ms 231/ 3341 (6.9%)
> >>>> [ 3] 5.0- 6.0 sec 1.59 MBytes 13.3 Mbits/sec 1.107 ms 16/ 1150 (1.4%)
> >>>> [ 3] 6.0- 7.0 sec 705 KBytes 5.77 Mbits/sec 25.142 ms 87/ 578 (15%)
> >>>> [ 3] 7.0- 8.0 sec 924 KBytes 7.57 Mbits/sec 0.454 ms 36/ 680 (5.3%)
> >>>> [ 3] 8.0- 9.0 sec 577 KBytes 4.73 Mbits/sec 15.732 ms 55/ 457 (12%)
> >>>> [ 3] 9.0-10.0 sec 1.81 MBytes 15.2 Mbits/sec 0.297 ms 58/ 1352 (4.3%)
> >>>> [ 3] 10.0-11.0 sec 728 KBytes 5.96 Mbits/sec 0.449 ms 0/ 507 (0%)
> >>>> [ 3] 11.0-12.0 sec 0.00 Bytes 0.00 bits/sec 0.449 ms 0/ 0 (nan%)
> >>>> [ 3] 12.0-13.0 sec 1.99 MBytes 16.7 Mbits/sec 0.545 ms 54/ 1472 (3.7%)
> >>>> [ 3] 13.0-14.0 sec 2.13 MBytes 17.9 Mbits/sec 0.321 ms 62/ 1583 (3.9%)
> >>>> [ 3] 14.0-15.0 sec 1.03 MBytes 8.64 Mbits/sec 0.716 ms 29/ 764 (3.8%)
> >>>> [ 3] 15.0-16.0 sec 2.14 MBytes 18.0 Mbits/sec 0.252 ms 214/ 1743 (12%)
> >>>> [ 3] 16.0-17.0 sec 2.96 MBytes 24.8 Mbits/sec 19.672 ms 61/ 2170 (2.8%)
> >>>> [ 3] 17.0-18.0 sec 40.2 KBytes 329 Kbits/sec 18.868 ms 16/ 44 (36%)
> >>>> [ 3] 18.0-19.0 sec 4.47 MBytes 37.5 Mbits/sec 0.185 ms 130/ 3316 (3.9%)
> >>>> [ 3] 19.0-20.0 sec 7.84 MBytes 65.7 Mbits/sec 0.210 ms 359/ 5949 (6%)
> >>>> [ 3] 20.0-21.0 sec 7.71 MBytes 64.7 Mbits/sec 0.149 ms 412/ 5914 (7%)
> >>>> [ 3] 21.0-22.0 sec 7.46 MBytes 62.6 Mbits/sec 0.674 ms 398/ 5721 (7%)
> >>>> [ 3] 22.0-23.0 sec 7.45 MBytes 62.5 Mbits/sec 0.725 ms 405/ 5719 (7.1%)
> >>>> [ 3] 23.0-24.0 sec 6.80 MBytes 57.0 Mbits/sec 0.374 ms 256/ 5106 (5%)
> >>>> [ 3] 24.0-25.0 sec 8.07 MBytes 67.7 Mbits/sec 0.163 ms 390/ 6145 (6.3%)
> >>>> [ 3] 25.0-26.0 sec 7.56 MBytes 63.4 Mbits/sec 0.230 ms 417/ 5809 (7.2%)
> >>>> [ 3] 26.0-27.0 sec 6.59 MBytes 55.3 Mbits/sec 1.414 ms 377/ 5080 (7.4%)
> >>>> [ 3] 27.0-28.0 sec 6.83 MBytes 57.3 Mbits/sec 0.224 ms 438/ 5308 (8.3%)
> >>>> [ 3] 28.0-29.0 sec 6.33 MBytes 53.1 Mbits/sec 7.292 ms 438/ 4954 (8.8%)
> >>>> [ 3] 0.0-29.7 sec 105 MBytes 29.7 Mbits/sec 1.346 ms 5453/80421 (6.8%)
> >>>> [ 3] 0.0-29.7 sec 1 datagrams received out-of-order
> >>>> _______________________________________________
> >>>> ath9k-devel mailing list
> >>>> ath9k-devel at lists.ath9k.org
> >>>> https://lists.ath9k.org/mailman/listinfo/ath9k-devel
> >>>>
> >>>
> >>
> >> yes it does at soem intervals as you can see.
> >> Have you solved or found the issue?
> >>
> >
>
> Does the link quality drop when the throughput goes down?
> I have -81 dBm at ~5 meters from the AP, no walls...
> Is the AR922X a single chip device as the AR928X?
AR9280 is the PCI express device, but PCI devices exist as well
and those get the AR922x name, there should be no difference
other than the bus used so yes these are single chip as well.
Luis
prev parent reply other threads:[~2010-01-04 15:55 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-12-24 13:27 [ath9k-devel] throughput is unstable rootkit85 at yahoo.it
2009-12-24 13:31 ` rootkit85 at yahoo.it
2009-12-24 18:19 ` Björn Smedman
2009-12-26 2:16 ` rootkit85 at yahoo.it
2009-12-26 9:27 ` Benoit PAPILLAULT
2009-12-26 12:38 ` rootkit85 at yahoo.it
2009-12-26 15:03 ` Benoit PAPILLAULT
2009-12-26 20:18 ` rootkit85 at yahoo.it
2009-12-29 0:24 ` Björn Smedman
2009-12-30 20:58 ` rootkit85 at yahoo.it
2010-01-04 15:55 ` Luis R. Rodriguez [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20100104155504.GC3855@tux \
--to=lrodriguez@atheros.com \
--cc=ath9k-devel@lists.ath9k.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.